HIV+ active and recovering drug users suffer from both poor utilization of, and low adherence to, primary care for HIV disease. Services that combine drug treatment and on-site primary care reduce these problems significantly because they create a support structure in which program staffs are able to monitor patients' health status, and give patients cues, reminders and reinforcement for keeping up with their treatments. However, only 10% to 15% of drug users are in drug treatment at any given time, and the vast majority of drug treatment programs do not provide primary care on-site. Thus only a small minority of drug users receive this demonstrably effective form of care. Hence, some alternative support structure is urgently needed for the large majority of HIV+ drug users who need help in obtaining and adhering to primary care services. This need is especially critical now that more effective antiretroviral therapy has become available with the advent of the protease inhibitors and potent combination regimens. Strict adherence to these regimens is essential, both for effectiveness and to prevent the emergence of HIV resistance. Studies of drug users' utilization of therapy - outside of structured settings like methadone treatment programs -- suggest that as a group they are less likely to benefit from these new therapies and more likely than other populations to spread multidrug-resistant HIV to others through unprotected sex and/or the sharing of syringes. This project's aim is to evaluate an innovative approach, termed a Peer-Driven Intervention, that harnesses peer support among drug users to promote admission into primary care, adherence to HIV therapeutics and reduction of HIV-related behaviors. The feasibility of the approach has been demonstrated by a multi-year, MDA-funded HIV-prevention project for injection drug users (RO1 DA08014), directed by the PI of this proposal. The proposed project also grows out of a pilot study that demonstrates the viability of the approach as a means for increasing adherence to medical care among HIV+ drug users. We propose to recruit and randomly assign 300 HIV+ active and recovering drug users to two treatment conditions, a Peer- Driven Intervention and a Usual-Care Intervention, and compare their enrollment and retention in primary care services, adherence to HIV therapeutics, entry and retention in drug treatment, reduction of HIV risk behaviors, and increase in medical knowledge. The study site, New Haven, Connecticut, is especially well suited for this community-based study. It is a geographically compact city with a highly integrated health care system that has been heavily impacted by HIV. The project brings together a strong interdisciplinary team with extensive experience in studies of drug users, AIDS, drug treatment, and innovative health and prevention services.